The goal of this application is to develop a very fast and cost-effective method to generate gene-targeted mice to model tumorigenesis. Gene-targeted mice are invaluable tools to determine the roles of oncogenic mutations in cancer development. However, conventional gene targeting is slow, expensive and prone to failure. While nuclease-mediated targeting may speed the production of mutants, there remain significant concerns about off-target mutations, relative ease of use and access to the entire genome. In preliminary studies, we have successfully developed an innovative method to directly and efficiently target mouse fertilized eggs using recombinant adeno-associated virus (rAAV)-mediated homologous recombination. Using this approach, we were able to generated germ-line-transmitting mice with at least 10% targeting frequency in a month. We believe that our technology is superior to nuclease-mediated gene targeting approaches (e.g. ZFN, TALEN and CRISPR/Cas). In contrast to nuclease-mediated approaches, off-target mutations are infrequent, embryos can be processed en masse without individual microinjection, and all regions of the genome are accessible to manipulation. Here we propose to further develop this technology to generate gene-targeted mice to model tumorigenesis by determining: (a) if gene-targeted mice generated by our method are suitable for modeling tumorigenesis, (b) if our approach is generally applicable to create gene-targeted mice of various tumor suppressors and oncogenes, and (c) if our approach can be used to generate conditional knock-out and knock-in mice. Successful development of these technologies will revolutionize generation of genetically engineered mice to model tumorigenesis. It will have huge impacts on basic cancer biology as well as cancer drug development.